Interface issues USG60

N77

My firewall performance drops over time and I can see increasing number of interface messages on syslog.

47 2018-08-21 19:49:36 debug Default kernel: eth0: 100 Mbps Full duplex, port 17       49 2018-08-21 19:49:27 debug Default kernel: eth0: 100 Mbps Full duplex, port 17       50 2018-08-21 19:46:37 debug Default kernel: eth0: 100 Mbps Full duplex, port 17

First this happens a few times every now (1-20 minutes) and then, after a few days I get these every few seconds. Reboot always helps for a couple days until throughput declines again.

I have changed all by cables and done the basics to figure out if this is external issue, but now I went on to try my secondary internet port for a day and I get no syslog errors on it at all.

According to graphs by observium the only notable difference is that eth_base acts differently while using wan1 and wan2 links, red shows my one day test on wan2 link



my config is fairly simple with only one minor differences; dhcp metric and ipv6 - ipv6 is globally disabled so I'd assume that has no impact as should not the dhcp metric.

interface-name ge1 Internet

interface-name ge2 Internet2

!

interface Internet

 ip address dhcp

 type external

 description Internet

 upstream 1048576

 downstream 1048576

 mtu 1500

 ipv6 dhcp6 address-request

 igmp activate

 no ip proxy-arp activate

!

interface Internet2

 type external

 description Internet2

 ip address dhcp metric 0

 upstream 1048576

 downstream 1048576

 mtu 1500

 igmp activate

 no ip proxy-arp activate

!

Any pro tips?

Thanks!

Zyxel_Stanley

Hi @N77

This log should coming from when physical layer is negotiating when linkup.

May we know more detail information on your environment:

(1) What is the speed supported on your device? 1000M or 100M? (maybe you can let us know the product name)

(2) Please use CLI command shows physical negotiation result when connecting WAN1 and WAN2 on USG

Router# show port status<b></b>

(3) How about the test result if connecting a switch between your modem and USG60?
e.g. Internet------modem-----switch------USG60

N77

Hi,

1)

I am not entirely sure what you mean by product name, other than USG60? 

Router# show version

Zyxel Communications Corp.

image number model                            firmware version                                                  build date           boot status

===============================================================================

1            USG60                            V4.32(AAKY.0)                                                     2018-07-12 10:31:26  Running

2            USG60                            V4.30(AAKY.0)                                                     2017-11-23 21:17:51  Standby

Link speed supported is 1000M but I have only 100M ethernet uplink.

2) 

Router# show port status

Port Status     TxPkts     RxPkts     Colli.  TxB/s      RxB/s      Up Time

===============================================================================

1    Down       10969779   12495178   0       0          0          00:00:00    

2    100M/Full  6188714    7546128    0       970        10984      03:50:11    

3    Down       0          0          0       0          0          00:00:00    

4    Down       0          0          0       0          0          00:00:00    

5    Down       0          0          0       0          0          00:00:00    

6    1000M/Full 20369606   17957030   0       12468      2222       36:20:56    

(I have switched to port 2 now as it seems to be more stable)

3)
I have ethernet link towards my ISP (shared fiber between apartments) and I am blindly connecting to their switch at 100F through ethernet jack in my apartment; ISP -> Shared switch -> USG60

However I have tried adding TP-link switch in between and the behavior is exactly the same.
ISP -> shared switch -> TP-link TL-SG2008 -> USG60 

Zyxel_Stanley

Hi @N77

Can you try to fix the port as full 100Mbps on your USG60.

 Please use this command:

Router# configure terminal

Router(config)# port status Port1

Router(config-port-status)# no negotiation

Router(config-port-status)# speed 100

Router(config-port-status)# duplex

Moreover, can you also check the port speed/duplex/auto negotiation modes on the TP-link switch settings? And also check the Ethernet cable is 8 pin.

N77

Hi,

Done but still get the message within minutes after completed

1

2018-08-23 15:21:05

debug

Default

kernel: eth0: 100 Mbps Full duplex, port 17

Also since I have TP link in between I can use 1000M link TP-link<-> USG60 and 100M ISP <-> TP so I tried both autoneg on TP-link and USG60 at 1000M and 100 full duplex without autoneg at both per your suggestion.

I am thinking at this point that the eth0 might actually just have hardware failure since everything works ok on eth1. (I see nothing physically wrong with eth0, no bent connectors or such)

I have cat6 and cat5 cables I tried with and the cable is 2 meters so I am gaining confidence in the fact that this is just a hardware failure at USG60 side.

What makes it odd is that reboot seems to help temporarily and it takes couple days before the performance goes really bad and the frequency of messages increases so I was thinking this could be some sort of counter overrun/leak issue too, but then again that should also impact second port?

Zyxel_Stanley

Hi @N77
I will send you private message for check this issue more detail.